Abstract: 【Objective】Casuarina equisetifolia is the main protection forest species in Hainan Island. The carbon sink capacity of this species has received considerable attention against the background of global climate change. 【Method】 In Daodong Forest Farm, Hainan Island, the 44 trees were selected in the C. equisetifolia forest as standard trees and their biomass was calculated. On the basis of the biomass data of the 41 selected trees and optimal models for above-and below-ground biomass, a compatible model for above-and below-ground biomass was established by using a non-linear error-in-variable modeling approach and conversion between variables of volume and above-and below-ground biomass.【Result】One variable equation based on diameter at breast height(DBH)was the optimal model for above-and below-ground biomass, and a binary equation based on DBH and height was the optimal model for tree volume. The use of the non-linear error-in-variable equation system was a solution to the problem of biomass compatibility, as the model R² values for above-ground biomass, below-ground biomass, and tree volume were all greater than 0.95, and this system also improved the prediction accuracy for individual trees(E_MPSE less than 10%)and managed to control E_MPE(mean prediction error). This study also developed a binary equation concerning the biomass conversion factor(E_BCEF)and root-shoot ratio(R). 【Conclusion】 The non-linear equation system provides a method for estimating the biomass and carbon storage of C. equisetifolia on a large scale.